HE decline of biodiversity is accelerated from the degradation, destruction and fragmentation of natural habitats. Field experiments show the consequences of fragmentation on both plants and animals. The ‘‘Biological Dynamics of Forest Fragments Project’’ (BDFFP) was an encompassing biodiversity study conducted in the Brazilian rain forests where stands of 1-, 10- and 100-hectare rain forest fragments were left intact and isolated from the clear-cut landscape. In a study of carrion and dung-feeding beetle communities, Klein (1989) showed that 1- and 10-hectare rain forest fragments, isolated from the contiguous forests by at least 350 meters for two to six years, had fewer species, sparser populations and smaller individuals than beetle communities in undisturbed rain forests. A study of avian species (Bierregaard & Lovejoy 1989) reported that as the rain forests were cut and cleared, birds moved into the remnant forest patches, initially raising the number of individuals. However, after 200 days the number of birds drastically dropped below pre-isolation numbers, with the rate and extent of faunal collapse being greater in smaller reserves than in larger reserves. The amount of time and effort necessary for an experiment like the BDFFP is extensive, so most experimental studies of fragmentation are limited to small habitat patches and attempts are made to extrapolate from them (Shafer 1990). On a smaller scale study, Robinson et al. (1992) recorded the responses of several plant and animal species to habitat fragmentation in pasture grasslands and noted that measures of species diversity did not change with the degree of fragmentation. Although, when looking within a single group, i.e. arthropods, large patches of uncut pasture (5,000 square meters or sq. m) contained a higher diversity of arthropod species (367 species) than medium-sized patches (288 sq. m; 315 species) and small patches (32 sq. m; 303 species).